Company
About Shieldra
Compliance software for companies that build or use AI. Shieldra AI, Inc. builds Shieldra.
Key takeaways
- Shieldra starts from the AI system — what it does and which role you hold — and derives obligations from published rules, each with its citation and deadline.
- 13 AI-law packs with 139 cited obligations, 6 compliance frameworks, and 125 rule test cases run in CI.
- Classifications come from a deterministic rules engine; language models help with drafting and document review but never decide a risk tier.
- Shieldra has not completed a third-party audit of its own, and says so on its Security and Trust Center pages.
Why Shieldra exists
Since 2 August 2026, Article 50 of the EU AI Act applies to products with a chatbot or AI-generated content, whatever their risk tier. Enterprise security questionnaires now include a section on AI governance. Most compliance tools treat AI as one more checklist.
Shieldra starts from the AI system: what it does and which role you hold for it. The obligations are derived from published rules, each with its article or section citation and deadline. SOC 2, HIPAA, HITRUST, NIST CSF, ISO/IEC 42001 and the NIST AI RMF live in the same workspace, so evidence is kept once.
How we work
- Regulation content is versioned data, not prose. Each pack records the date it was last checked against the official text.
- Classifications come from a deterministic rules engine. Language models help with drafting and document review; they never decide a risk tier.
- We state our own security posture plainly, including what has not been audited yet.
- Shieldra is not a law firm. Confirm scope decisions with counsel.
Company
- Legal entity: Shieldra AI, Inc.
- Support: support@shieldra.ai
- Security: security@shieldra.ai
- Privacy: privacy@shieldra.ai
- Legal: legal@shieldra.ai
Frequently asked questions
Who is Shieldra built for?
Companies that build or use AI — especially B2B software teams selling into the EU or to enterprise buyers — that need AI governance and security compliance without a dedicated compliance team.
How does Shieldra decide an EU AI Act risk tier?
A deterministic rules engine evaluates your answers against a versioned content pack: a scope rule, then tier rules in a fixed order, then the obligations for your role, each with its citation and deadline. The rule trace is saved. No language model decides the outcome.
Does Shieldra store protected health information?
The standard platform is a No-PHI service intended for compliance documentation, policies, risk assessments, and audit evidence. PHI processing requires separate written authorization under a Business Associate Agreement.